Browsing by Subject "missing data"

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Open Access
    Making sense of censuses and dealing with missing data: trends in pup counts of Cape Fur Seals Arctocephalus pusillus pusillus between 1972-2004
    (2007) Kirkman, S P; Oosthuizen W H; Meÿer, M A; Kotze, P G H; Roux, J-P; Underhill, L G
    Trends in the population of Cape fur seals Arctocephalus pusillus pusillus were estimated from counts of pups on aerial photographs of colonies taken between 1972 and 2004 to determine trends in the overall population and subpopulations. Incomplete coverage resulted in missing data in some years. Various methods of determining proxy values for missing data were assessed, and it was concluded that different methods were applicable to Namibian and South African colonies. This reflected variation in trends of pup counts between the countries, which was associated with differences in productivity between the southern and northern Benguela ecosystems. In Namibia, temporal changes in pup numbers were non-linear in some years and there was correspondence in fluctuations at most colonies. This appeared to be on account of an effect of periodic, wide-scale prey shortages that reduced birth rates. There was a northward shift in the distribution of seals in the northern Benguela system. In South Africa, pup counts were much less variable between years, probably on account of a relative stability of food supply. A linear approach was therefore suitable for determining proxy values for missing data at South African colonies. Pup counts suggest that there has been little change in the overall population of the Cape fur seals since 1993, when it was estimated at about two million animals.
  • Thumbnail Image
    Item
    Metadata only
    MAMI: An R-package for model selection and model averaging after multiple imputation
    (2015-09-07) Schomaker, Michael
    The package performs model selection/averaging on multiply imputed datasets and combines the resulting estimates. The package also provides access to less frequently used model averaging techniques and offers integrated bootstrap estimation.